Season 3, Episode 7 (Episode #43)
Solo episode. No guest. Doug introduces a three-layer framework for AI output quality — Intent, Context, and Prompt Engineering — and explains why most designers are over-investing in the least important layer. Uses two fictional designer archetypes to illustrate both ends of the experience spectrum. Introduces his Socratic AI tools as the solution to the context extraction problem that the framework alone doesn't solve.
Chapter Timestamps
[00:00] Why your AI results are "fine but not good enough"
[00:41] Meet Maya — the tech-comfortable designer still getting mediocre results
[03:40] The three layers that determine AI output quality
[04:05] Layer 1: Prompt engineering — and why it's the least important
[05:23] Layer 2: Context engineering — what AI doesn't know about you
[06:13] The junior designer analogy
[07:29] Layer 3: Intent engineering — outcome vs. task
[08:43] How the three layers work together (the hierarchy)
[09:02] The outreach email example — generic vs. intentional
[11:12] Meet Carol — 22-year veteran who set AI aside
[13:16] Why experienced designers can't manually transfer expertise to AI
[13:33] Doug's Socratic AI tools — drawing context out through questions
[14:46] Carol's breakthrough: 22 years of expertise finally articulated
[18:14] What's coming next + how to work with Doug
More videos on intent and context engineering: coming soon
